Ph.D.
Browse
4 results
Search Results
Item Heuristic Approaches for Maximin Distance and Packing Problems(Università degli Studi di Torino, Torino, Italy, 2009-02-18) Jamali, Abdur Rakib Muhammad Jalal Uddin; Locatelli, Prof. MarcoIn this thesis we mainly deal with two problems - experimental design and packing problems. In the field of experimental design problems we consider maximin Latin Hypercube Designs (LHDs). In the field of packing problems we consider those of packing n equal or unequal circles in a circular container with minimum radius. Both problems can be formulated as optimization ones. The former is a combinatorial problem, while the latter is a continuous one. We propose heuristic approaches to tackle these problems. These are Iterated Local Search (ILS) heuristics for maximin LHDs, and Basin Hopping (BH) heuristics for packing problems. Actually, ILS and BH approaches have strong similarities and could be described within an unified framework. However, following the literature, where ILS approaches are mainly applied to combinatorial problems, while BH approaches are mainly applied to continuous problems, we will keep them apart. In order to deal with maximin LHDs, we propose two ILS variants, corresponding to two distinct optimality criteria which are employed to drive the search among LHDs. Extensive experiments are performed for the investigation of the strengths and weaknesses of the algorithms. A remarkable finding is that the most efficient method, though time consuming, performs a non monotonic search, driven by an appropriate objective function, within the space of LHDs. The proposed approaches are extensively compared with the existing ones in the literature, and many improved results with respect to best known ones, are obtained. In particular, the proposed methods seem to outperform the existing ones when the dimension of the design points increases. Finally, we also discuss about the time complexity of the algorithms; by mixing theoretical results with experimental ones, we derive an empirical formula for each ILS variant, returning the expected run time as a function of the number of design points and of their dimension. To deal with the problem of packing equal circles in a circular container with minimum radius, we propose a variant of BH, namely Monotonic BH (MBH) and its population based counterpart, Population BH (PBH). Extensive computational experiments are performed both to analyze the problem at hand, and to choose in an appropriate way the parameter values for the proposed methods. Different improvements with respect to the best results reported in the literature are detected. The problem of packing unequal circles in a circular container with minimum radius is also attacked with the MBH and PBH approaches, but some components of these approaches are adapted in order to fully exploit the peculiarities of the problem with unequal circles (in particular, its combinatorial nature due to the different radii of the circles). Again extensive computational experiments are performed and improvements with respect to the existing literature are detected.Item Planar Electromagnetic Bandgap Structures and Applications(Nanyang Technological University, Nanyang Avenue, Singapore, 2005-03) Mollah, Md. Nurunnabi; Fu., Dr. Jeffrey S.With the emergence of novel wireless technologies, the demands for larger bandwidth, high efficiency and high capacity RF and microwave devices have been growing tremendously. To satisfy the demand novel materials, new processing techniques and refinement of existing devices have been evolving. Using electromagnetic bandgap (EBG) is a new technology to improve the performances of existing RF active and passive devices. EBG structures (EBGSs) are periodic structures, which exhibit distinct passband and stopband characteristics. The passband can be used as phase shifters and slow wave structures. The stopband characteristics can be used to suppress surface waves in dielectric media. Due to these unique properties, EBGSs find potential applications in many active and passive RF and microwave devices including antennas, filters, amplifiers and oscillators. Recently, the dispersion properties in passband of EBGSs are used in the phased array antennas for wide angle beam steering. Among various EBG configurations, planar EBGSs become most popular due to their low profile, ease of fabrication and integration with monolithic microwave integrated circuits (MMICs). The thesis concerns the planar EBG structures in the forms of conventional circular and rectangular photonic bandgap structures (PBGSs) and defected ground structures (DGSs). Novel PBGSs in the form of non-uniform Binomial and Chebyshev distributions of unit PBG cells have been proposed. This novel PBGSs yield ripple free passband and wide stopband properties, which are very useful to suppress higher order harmonics in the bandpass filters. Next the chirped PBGSs are investigated that yield better passband return loss, low ripples and wider stopband than those for conventional PBGSs as well as non-uniform PBGSs mentioned earlier. The parametric study of dumbbell shaped DGSs has been conducted with the gap width and length and the size of dumbbell slots. Three novel designs are proposed: non-uniform dumbbell shaped DGSs with Chebyshev distribution, hybrid DGS and PBG and modified hybrid DGS and PBG configurations. The novel designs outperform the conventional DGSs reported in the literature. The DGSs yield perfect LPF responses with negligible passband ripples and extremely wide stopband. The modified hybrid DGS and PBG configuration yields dual stopbands that can be used as a dual stopband filter. The novel design are used in asymmetric coupled line bandpass filters, aperture coupled single band and dual-band patch antennas and finally in 4-element conventional and reconfigurable phased array antennas. For filters, PBGSs suppress second and third harmonics by about 40 dB, A comprehensive investigation of the number, filling factor and position of the PBGS under the filter has been conducted. The frequency response of the filter has been observed. This investigation is novel and illustrates the significance of PBGSs in suppression of higher order harmonics of the bandpass filters. It has been found that non-uniform distribution of PBGSs is more effective to suppress higher order harmonics than the conventional uniform PBGSs. DGSs are very effective for simultaneous suppression of both 2nd and 3rd harmonics due to their very wide stopband performances. This phenomenon of DGSs has been demonstrated in various designs of asymmetric coupled line bandpass filters. Finally, the passband phase properties of the PBG and DGS have been investigated. The comprehensive investigations of relative phase delays of PBGS and DGS assisted microstrip transmission lines have been performed. Such investigation is not found in the open literature. It is observed that DGSs yield much larger phase delay compared to a similar size PBG structure. Therefore, DGS assisted feed network can scan the beam over a wide angle. After thorough and satisfactory investigation of phase delays for PBGSs and DGSs, phased array theory is studied. The required phase delays for a 4-element phased array are calculated based on the element spacing and number of elements. A wide beam scanning up to approximately 60° is achieved with 0-8-16-24 DGS assisted corporate feed network for the 4-element patch antenna array. So far such wide angle scanning using DGS assisted beamforming network has not been reported in the open literature.Item Stress-Deformation Characteristics of Selected Coastal Soils of Bangladesh and their Sampling Effects(Bangladesh University of Engineering & Technology (BUET), Dhaka, Bangladesh, 2002-03) Bashar, Md. Abul; Safiullah, Prof. Dr. A.M.M.A considerable development activity within the Coastal Region of Bangladesh has necessitated an understanding of the geotechnical behaviour of soils from this region. With this objective in view a study into strength-deformation, compressibility and intrinsic properties of reconstituted samples of Chittagong coastal soils were undertaken. This thesis presents stress-deformation characteristics of three selected coastal soils and their sampling effects. The soils were collected from Banskhali, Anwara and Chandanaish in Chittagong coastal belt of Bangladesh. The soils are low to medium plasticity (Liquid limit = 34 to 45 and Plasticity index = 10 to 20). Reconstituted samples of the three soils were prepared in the laboratory by K0- consolidation of slurry in a large cylindrical consolidation cell using a consolidation pressure of 150 kN/m2. Overconsolidated samples were prepared in the triaxial cell by releasing the maximum isotropic consolidation pressure of 150 kN/m2 to appropriate values to achieve overconsolidation ratios (OCR) of 1.5, 2, 5, 10, 20 and 30. The stress-deformation-strength, stiffness and pore pressure characteristics of reconstituted isotropically normally consolidated and overconsolidated "block" samples of the three coastal soils were investigated in the laboratory by performing undrained triaxial compression tests. Models for the prediction of undrained shear strength of normally consolidated and overconsolidated samples have been developed. To develop intrinsic models of compressibility, intrinsic compression lines (ICL) for the three soils under K0 and isotropic stress conditions have been established which can be used to determine compressibility indices of soils at any depth of known overburden pressure. State boundary surfaces (Roscoe and Hvorslev state boundary surfaces) and critical state lines of the three coastal soils have been established. The critical state parameters of the soils have also been evaluated. Constitutive models relating critical state soil parameters and plasticity index of the soils have been proposed. Applications of these models to undisturbed naturaj clays will require further investigation. The experimentally observed stress-strain behaviour of reconstituted normally consolidated samples of the three coastal soils have been compared with those predicted using two critical state models, namely, "Cam clay model" and "Modified Cam clay model". It has been found that the results predicted by using "Modified Cam clay model" compared more favourably with the observed experimental results than "Cam clay model" for the three coastal soils. The present study has also been carried out to investigate the effects of "perfect" sampling disturbance and tube sampling disturbances on engineering properties of reconstituted normally consolidated samples of the three coastal soils. Undrained triaxial compression tests were carried out on "in situ", "perfect" and "tube" samples. "In situ" samples were prepared by consolidating reconstituted specimens of 38 mm diameter by 76 mm high under K0-condition in the triaxial cell to its in situ stress state. "Perfect" samples were prepared from "in situ" samples by undrained release of the total stresses in the triaxial cell. "Tube" samples were prepared from the large diameter consolidated samples by inserting samplers of different area ratios, external diameter to thickness ratio (De/t) but of constant outside cutting edge angle (OCA) and internal diameter (Di). Area ratio, De/t ratio, OCA and Di of the samplers were 16.4% to 73.1%, 27.3 to 8.3, 5° and 38 mm, respectively. Undrained triaxial compression tests were carried out on reconsolidated "perfect" and "tube" samples of the three coastal soils to assess the suitability of various reconsolidation techniques to minimize sampling disturbance effects. Experimental results indicate that disturbances due to perfect and tube sampling have significant influence on the mechanical properties of coastal soils. The nature of the effective stress paths and por' pressure responses of both "perfect" and "tube" samples are markedly different from those of the "in situ" samples. The "perfect" and "tube" samples adopted stress paths and showed pore pressure responses which are more typical of overconsolidated clays. Disturbances due to perfect sampling led to reduction in the values of undrained shear strength (se), Skempton's pore pressure parameter A at peak deviator stress (Ap), initial tangent modulus (Ei) and secant stiffiess at half the peak deviator stress (E50) whilr axial strain at peak deviator stress (εp) increased due to total stress relief. Due to total stress relief, the reduction Su, Ei and E50 increased with the decrease of plasticity while the increase in c. increased with the decrease of plasticity of the soils. It is also evident that the decrease in mean effective stress (p') due to perfect sampling increases with decreasing plasticity of the soils. The initial effective stress (σ՛i) of "tube" samples reduced considerably because of disturbance caused by penetration of tubes. Compared with the "in situ" samples, values of s, E1, E50 and A of the "tube" samples decreased while sp increased. The changes in measured soil parameters between the "in situ" and "tube" san )les have been found to depend significantly on the sampler characteristics, i.e., area ratio, De/t ratio, used for retrieving the "tube" samples. The values of σ՛i, su, Ei and E50 were decreased due to increase in area ratio (or reduction in De/t ratio). The values of εp, however, increased due to increasing area ratio. A quantitative increase in the degree of disturbance (Dd) has been obtained due to increase in area ratio, while the values of Dd increased with the decrease of De/t ratio of sampler. Disturbance due to tube sampling has been found to depend on the plasticity of the samples of the three coastal soils. The highest reductions in a', st,. Ei and E 0 occurred in the least plastic samples, whereas the minimum reduction in σ՛i, su, Ei and E50 occurred in the most plastic samples. Among the samples of the coastal soils, the least plastic sample produced higher degree of disturbance than the most plastic sample.Item Investigation of Strength and Deformation Characteristics of Cement and Lime Treated Soft Clays(Bangladesh University of Engineering & Technology (BUET), Dhaka, Bangladesh, 2008-07) Rahman, Md. Mokhlesur; Siddique, Prof. Dr. AbuThis thesis presents experimental and numerical investigations for the effects of cement and lime treatment on compressibility, permeability, stress-strain, strength and stiffness behaviour of three soft clays. Effects of cement treatment on physiochemical and micro-structural properties have also been investigated. X-ray Diffraction, Scanning Electron Microscopy, particle size distribution, pH measurement, organic content, electrical conductivity, cation exchange capacity, exchangeable cation, water content, unit weight, specific gravity, and Atterberg limits tests were conducted to examine physiochemical and micro-structural properties. Compressibility and permeability properties of untreatcd and, cement and lime treated clays were investigated by performing one-dimensional consolidation tests. Stressstrain, strength and stiffness behaviour of untreated and, cement and lime treated clays were evaluated by performing unconfined compression (UC) tests. Consolidated drained direct shear (OS), unconsolidated undrained (UU), isotropically consolidated undrained (CIU) and isotropically consolidated drained (CID) triaxial compression tests were also carried out to assess the stress, deformation and strength properties of cement treated clays. Sets of variables considered in the testing program include a wide range for type of clay (PI = 13% to 47%), type of admixture, clay-water/cement ratio (2 to 30), curing time (1 to 104 weeks), mixing water content (120% to 250%) and effective confining pressure (50 kPa to 400 kPa). Finally, applicability of different constitutive models and Cap models were assessed for the predictions of drained and undrained behaviour. It has been observed that pH value, electrical conductivity (EC), cation exchange capacity (CEC), increases with decreasing clay-water/cement ratio. pH value, however, decreases while EC and CEC value increases with increasing curing time. Loss on ignition and organic matter decreases due to increasing cement content and with increasing curing time. The relative amount of cementitious product (CSH + CASH) was identified by the XRD analysis and found to increase with the increase of cement content and curing time. Due to the formation of cementitious product, the fabric of the treated clays changed to flocculated type, comprising of clay-cement clusters separated by large inter-cluster voids with smaller intracluster pores as seen from the SEM images of treated clays. These changes were more pronounced with higher cement content and longer curing time. The significant increase in yield stress and reduction in compression index (Ce) and swell index (Cs) were observed with increasing admixture (cement/lime) content and increasing curing time. Ce and Cs for lime treated clays were found to be greater than those of cement treated clays. Ce and Cs values increase significantly with the increase of mixing clay-water content. The lime-treated clays gained comparatively higher void ratio and volumetric strains and lower yield stress than those of cement-treated clays. The effect of cementation is to increase the values of coefficient of consolidation (cv) and coefficient of volume compressibility (mv). The higher the admixture content and curing time, the lower is the values of cv and mv. cv and mv values for lime treated clay were found to be higher than those of cement treated clay. A substantial increase in the value of cv and mv was found to have occurred within higher mixing water content. At higher stress level, progressive destructuration of the treated clay particles occurred, which has been verified from the SEM images of cement treated clay compressed at different consolidation pressure. Addition of admixture to the clay increases the permeability and void of the soils, due to flocculation of the soil particles, which has been seen from SEM images. The Intrinsic Compression Line (ICL) and Generalized Compression Line (GCL) have been proposed for the untreated and treated clays. Coefficient of permeability (k) and void ratio relationships have also been proposed for cement and lime treated clays. The values of k of cement and lime treated clays have been found to reduce with increasing cement or lime content and curing period. Values of k for lime treated clays were found to be higher than those of cement treated clays. Based on the rate of strength development with time, the unconfined compressive strength and cement content relationships have been divided into 3 zones: Inactive Zone, Active Zone and Inert Zone. Since the behaviour of cement treated clays was remarkably governed by we/e, the strength prediction in terms of we/e as well as the interrelationship involving strength, curing time and clay-water content/cement content ratio (we/e) have been proposed. From stress-strain relationships, the overall behaviour has been categorized into brittle, quasi-brittle and ductile. Comparatively, brittle, quasi-brittle and ductile types for the cement treated clays, while quasi-brittle and ductile types for the lime treated clays have been found. The correlation between yield stress, (σy') and unconfined compressive strength (qu), τf0 and qu axial strain at failure (εf) and qu, and stiffness (Ei and E50) and qu have also been proposed. In direct shear test, for cement treated clays, vertical expansion (dilation) was observed at low normal stress and experienced vertical contractions or settlement throughout the shearing stage at higher normal stress. The cohesion and friction angle have been increased with increasing cement content and curing time. The undrained effective stress paths of the cement treated clays obtained from CIU triaxial compression tests indicate that the stress paths belong to different category of states such as normally consolidated, lightly, moderately and heavily over-consolidated state. The degree of alteration have been found different for different samples depending on the amount of cement content, curing time and pre-shear effective consolidation pressure. Upon reaching the peak deviator stress in CIU triaxial test, the progressive destructuration takes place and thus the stress path tends to move either on the I-Ivorslev envelope or envelope of strain softening behaviour. SEM results have been suggested that complete destructuration takes place only on the shear plane at which the clay-cement cluster crushes. The prevalent role of pre-shear effective consolidation pressure was manifested to annihilate the cementation effect attributing ductility to the treated matrix. The main effect of cement treatment was to modify the behaviour of the soft clay form normally consolidated to over-consolidated state. Similar natures of deviator stress-axial strain and volumetric strain-axial sirain relationships were observed for all samples having the identical wclc ratio in cm triaxial tests. So, the wclc is a prime parameter governing the engineering behaviour of cement treated samples having different mixing water content. For cement treated clays, peak deviator stress (qmax) criterion of failure envelopes can be used as obtained from CIU and CID triaxial compression tests. The degree of overall curvature of the failure envelope for each type of clay depends on the range of consolidation stresses and hence increases with increasing cement content and curing time. The values of soil constants λ and κ for the treated clays have been found to be less than those for the untreated clays while the values of the constant N and M are greater for the cement treated clays than those of untreated clays. At a particular curing time, the soil constants λ and κ decreased, while the constant N and M increased with increasing cement content. Correlations between the soil constants λ, κ and N with plasticity index have been proposed. Hvorslev surface was established for the cemented clays. No definite Roscoe surface could be found for cemented clays. For the untreated clay, it has been observed that the predicted stress paths, deviator stresses, volumetric strains and excess pore pressure responses at small strain levels using the MCC model appear to be significantly close to experimental curves. It appears from the present study that MCC, MMCC, EMMCC and Cap models (Plane Cap and Elliptic Cap) cannot be applied for predictions of drained and undrained behaviour of cemented clays at high water content.